tacit · Insurance
tacit
Organisational Intelligence Infrastructure
35% of senior underwriters retiring in five years. Their judgment isn't in any system. We capture it before it leaves.
The Problem Nobody Talks About
88% of enterprises have adopted AI. Only 5% see real returns.
— BCG + McKinsey, 2025
85% of AI projects never reach production.
— Gartner, 2025
The difference isn't technology. Both groups use the same models.
The difference is organisational.
— Deloitte State of AI, 2025
The Missing Layer
Your best underwriter looks at a submission and sees risk nobody else catches.
Your senior buyer knows which supplier will deliver late before the data does.
Your compliance lead knows which regulatory nuance will sink a deal.
None of this is written down. It lives in their heads — heuristics built
over decades. When they leave, their expertise leaves with them.
Your AI doesn't fail because the models are bad.
It fails because it doesn't know what your best people know.
The Knowledge Hierarchy
Organisational Intelligence is all three layers combined.
Most systems only reach the top. We capture what's underneath.
What Tacit Builds
Behavioural Analytics Engine
—
We instrument your existing platforms and observe how your experts actually
make decisions. What data they check, in what order, what they skip. Continuous,
passive, zero-intrusion — like UEBA, but for expertise instead of threats.
Knowledge Graph
—
We capture your organisation's tacit knowledge — the unwritten heuristics,
decision patterns, and exception-handling logic — and encode it into a
queryable graph that AI systems can traverse. Not documents. Decision intelligence.
Intent Engine
—
We encode what your organisation actually wants to achieve — objectives,
constraints, trade-off preferences — as machine-readable parameters. Your AI
stops optimising for the wrong thing.
Alignment Monitor
—
Drift detection, decision logging, knowledge freshness scoring, and ROI
measurement. Your AI stays aligned as your organisation evolves. When it
diverges, you know immediately.
How It Works
Week 1–8: Discovery
—
Shadow sessions with your best performers. Decision archaeology on
50–100 past decisions. Scenario elicitation for edge cases.
You get a deployment-ready intent specification and knowledge graph.
Month 2–4: Infrastructure
—
We deploy the behavioural analytics SDK on your platforms, the knowledge
engine, and the intent alignment layer. Your AI starts thinking
like your best people — not a generic model.
Ongoing: Operations
—
Passive capture enriches the knowledge graph continuously.
When experts change their approach, we detect it. When someone
retires, their expertise stays. The system improves as your organisation evolves.
Why Now
The workforce cliff is here
—
10,000 baby boomers retire every day across the US and UK.
Insurance underwriting, procurement, and compliance are dominated
by experts aged 50+. Their knowledge is leaving faster than it can
be transferred. Traditional training takes 5–7 years to produce a
competent underwriter. The pipeline is empty.
LLMs made extraction possible
—
Before 2023, encoding unstructured expert reasoning into machine-readable
form was a research problem. Foundation models can now parse natural
language heuristics, classify decision patterns, and generate structured
knowledge representations. The core technology bottleneck has lifted.
Regulators are demanding explainability
—
The EU AI Act (2025), PRA SS1/23 model risk guidance, FCA Consumer
Duty, and Lloyd's AI governance requirements all mandate that
organisations explain how AI-assisted decisions are made. If your experts'
decision logic isn't documented, you have a compliance gap today.
Three forces converging: expertise is leaving, extraction is now
technically possible, and regulators demand it. The window is 18–36 months
before the knowledge is gone permanently.
The Ground Reality
2025
82% of insurance leaders prioritise AI adoption.
Only 22% have deployed at scale.
—
Fintech Global
2025
Winners focus on 3.5 use cases. Losers scatter across 6.1
—
BCG
2025
30% of GenAI projects abandoned after proof-of-concept
—
Gartner
2025
84% of enterprises haven't redesigned jobs for AI.
They layer AI on existing processes and wonder why it doesn't work.
—
Deloitte
2025
AI underwriting automation market: $410M today. $7.9B by 2033.
44.7% CAGR.
—
Congruence Market Insights
Who This Is For
Your expertise is walking out the door
—
Senior underwriters retiring. Veteran procurement leads moving on.
Compliance specialists who've seen every edge case — leaving
and taking 30 years of unwritten knowledge with them.
Your AI is fast but not smart
—
You've automated processing speed. Document extraction works.
But the quality of judgment hasn't improved because your AI
doesn't know what your best people know. Speed without wisdom.
Your regulators want to see your thinking
—
Solvency II, IFRS 17, emerging AI regulation — compliance
requires you to explain how risk decisions are made. If your best
people's decision logic isn't documented, you have a governance gap.
A New Category
Knowledge management captures what people write down.
RAG retrieves existing documents.
Process mining shows what happens, not why.
Nobody captures how your best people actually think — the
heuristics, the judgment calls, the pattern recognition built
over decades — and encodes it for AI.
We're building Organisational Intelligence Infrastructure.
The layer between AI models and enterprise reality.
What We Are — What We Are Not
Not Knowledge Management
—
Confluence and SharePoint capture what people write down.
We capture what they can't articulate. If your best underwriter
could put everything she knows into a wiki, you wouldn't need us.
She can't. That's the point.
Not RAG
—
Retrieval-augmented generation retrieves existing documents.
The most valuable knowledge in your organisation has never been
written down. RAG retrieves nothing. We create the knowledge
that RAG should be retrieving.
Not Process Mining
—
Celonis tells you that your best underwriter takes three extra steps.
We tell you why she takes those steps — and why those steps are
the reason she's your best underwriter.
Not Decision Intelligence
—
Palantir builds intelligence from data. We build intelligence
from expertise. They start top-down with transactions. We start
bottom-up with people. Complementary — not competitive.
Before you can manage knowledge, you have to capture it.
Before RAG can retrieve expert judgment, someone has to extract it.
We're the layer before the layer.
Security & Compliance
Data residency
—
All data processed and stored within EU/UK jurisdiction.
No cross-border transfer. Your data never leaves your regulatory zone.
Encryption
—
AES-256-GCM at rest. TLS 1.3 in transit. Zero plaintext
storage of behavioural data or extracted heuristics.
Compliance readiness
—
SOC 2 Type II in progress. GDPR compliant by design — behavioural
analytics captures decision patterns, not personal data.
Supports PRA SS1/23, FCA Consumer Duty, EU AI Act Article 14
human oversight requirements.
Certifications
—
Cyber Essentials certified. ISO 27001 roadmap in progress.
Annual third-party penetration testing. Template DPA
available on request for accelerated procurement.
Insurance & liability
—
£2M Professional Indemnity insurance. £5M Public Liability.
Vendor security questionnaire pre-completed for major
insurance platforms.
Deployment model
—
Runs within your cloud tenancy or as a managed service.
No data leaves your perimeter. Read-only SDK integrates
with Guidewire, Duck Creek, Sapiens, and custom platforms.
No PII in the knowledge graph — we capture decision patterns,
not personal data.
Early Signal
Working prototype capturing expert decision sequences
from browser-based underwriting platforms.
Conversations active with London Market syndicates
and UK specialty insurers for Phase 1 deployment.
Methodology validated through structured expert interviews
with senior underwriters (25+ years experience) across
property, casualty, and specialty lines.
Technical architecture supports 10,000+ events/second ingestion
with sub-100ms latency — built for enterprise-grade throughput
from day one.
Who We Are
Jainam Shah — Founder & CEO
—
Enterprise AI and insurance technology. Built knowledge
systems for underwriting teams at global insurers.
Understands both the technology stack and the operational
reality it serves.
LinkedIn ↗
Advisory
—
Insurance domain from senior practitioners across Lloyd's,
London Market, and global reinsurance. Technical guidance from
AI/ML researchers with published work in knowledge representation
and behavioural analytics.
We're not building AI. We're building the infrastructure
that makes your AI understand your business.
Start a Conversation
Week 1: Discovery call
—
30 minutes. We assess fit — your decision complexity,
expert concentration risk, and AI maturity. No pitch deck.
Mutual evaluation.
Week 2–3: Diagnostic
—
We analyse 3–5 expert decision workflows. You get a written
assessment of your tacit knowledge concentration risk and
a preliminary heuristic map — yours to keep, even if you don't proceed.
Week 4–8: Proof of value
—
Full shadow sessions with your top performers. Deployed
knowledge graph with extracted heuristics. Measurable before/after
on decision quality and consistency.
Capacity-limited by design. We partner deeply with 3–5 enterprises
at a time. No surface-level deployments.